Throughput accounting asks managers to stop improving every resource independently and focus on the one resource that limits the performance of the system as a whole: the bottleneck.
In a short-run throughput model, direct material is treated as the only truly variable cost. Labour and other operating costs are treated as fixed. The objective is therefore to maximise throughput generated from sales while managing the constraint and total factory cost.
Current PM scope: the September 2026 to June 2027 syllabus requires the theory of constraints, the throughput accounting ratio (TPAR), ways to improve TPAR and multi-product decisions. Backflush accounting appeared in the old F5 download but is not a listed current PM learning outcome, so it is deliberately omitted here.
1. Start with the goal and the constraint
A production system is a chain of connected activities. Its output cannot exceed the capacity of its weakest link. If one machine can process only 500 units per week while every other stage can process 800, improving a non-bottleneck from 800 to 900 does not increase finished sales. It may simply create more work in progress waiting in front of the bottleneck.
Theory of constraints therefore evaluates the whole system, not isolated department utilisation. The aim in a profit-seeking organisation is to increase the rate at which the system generates money through sales.
2. The throughput accounting assumptions
Direct material is the only totally variable cost in the short-run model.
Direct labour and other factory costs are treated as fixed operating expenses over the relevant period.
Throughput is generated only when a unit is sold. Producing inventory does not create throughput.
There is at least one constraint that limits total output.
Management should maximise throughput per unit of the bottleneck resource, not contribution per unit of product.
These are simplifying assumptions. In a real organisation, some non-material costs may be variable or avoidable. In the exam, follow the facts given and state any limitation relevant to the decision.
3. The four calculations you must know
1. Throughput per unit = selling price − direct material cost
2. Return per factory hour = throughput per unit ÷ bottleneck hours per unit
3. Cost per factory hour = total factory cost ÷ total bottleneck hours available
4. TPAR = return per factory hour ÷ cost per factory hour
“Factory hour” means an hour of the bottleneck resource. If quality-control time is the constraint, the return must be calculated per quality-control hour, not automatically per machine hour.
TPAR result | Interpretation |
|---|---|
Above 1 | The product generates throughput faster than the system incurs factory cost per bottleneck hour. |
Equal to 1 | The throughput rate equals the factory-cost rate. |
Below 1 | The throughput rate is below the factory-cost rate; management should investigate price, material cost, bottleneck time and capacity. |
Do not use TPAR mechanically as a product-abandonment rule. If factory costs are genuinely unavoidable and otherwise-idle bottleneck capacity exists, a sale with positive throughput may still improve total profit. If scarce capacity can be used for a higher-return product, it should be prioritised there.
4. The five focusing steps
Identify the constraint. Find the resource whose capacity prevents the system meeting demand.
Exploit the constraint. Use existing bottleneck capacity as effectively as possible: avoid idle time, breakdowns, defects and low-value work.
Subordinate everything else. Schedule non-bottlenecks to support the constraint. Do not produce work merely to keep every department busy.
Elevate the constraint. If justified, add capacity through investment, overtime, subcontracting, process redesign or additional staff.
Repeat. Once one constraint is relieved, another part of the system becomes the new bottleneck.
Study rule: improving a non-bottleneck does not increase throughput unless it helps the bottleneck or prevents the non-bottleneck becoming the next constraint.
5. Worked example: identify the bottleneck
A highly automated factory makes three products. All costs except direct material are treated as fixed for the period.
Per unit | Product A | Product B | Product C |
|---|---|---|---|
Maximum demand (units) | 8,000 | 10,000 | 6,000 |
Selling price | $130 | $100 | $135 |
Direct material | $33 | $20 | $40 |
Machine hours | 0.25 | 0.20 | 0.30 |
Labour hours | 0.25 | 0.20 | 0.30 |
Quality-control hours | 0.10 | 0.10 | 0.10 |
Available capacity is 5,000 machine hours, 6,000 labour hours and 2,500 quality-control hours.
Resource | Hours needed for maximum demand | Hours available | Conclusion |
|---|---|---|---|
Machine | (8,000 × 0.25) + (10,000 × 0.20) + (6,000 × 0.30) = 5,800 | 5,000 | Constraint |
Labour | 5,800 | 6,000 | 200 hours spare |
Quality control | (8,000 + 10,000 + 6,000) × 0.10 = 2,400 | 2,500 | 100 hours spare |
Machine time is the bottleneck because demand requires more machine hours than are available.
6. Rank products by return per bottleneck hour
Measure | Product A | Product B | Product C |
|---|---|---|---|
Throughput per unit | $130 − $33 = $97 | $100 − $20 = $80 | $135 − $40 = $95 |
Machine hours per unit | 0.25 | 0.20 | 0.30 |
Return per machine hour | $388.00 | $400.00 | $316.67 |
Priority | 2 | 1 | 3 |
Product A has the highest throughput per unit, but Product B uses scarce machine time more efficiently. Product B must therefore be produced first.
Try the ranking calculation in the spreadsheet
Enter formulas in the yellow cells. Use the $324.80 cost per factory hour when calculating TPAR, then select Show answer to compare your work. Which product should be ranked first?
| Product | Selling price | Direct material | Bottleneck minutes | Throughput/unit | Return/hour | TPAR |
| A | 130.00 | 33.00 | 15.00 | |||
| B | 100.00 | 20.00 | 12.00 | |||
| C | 135.00 | 40.00 | 18.00 |
7. Build the optimum production plan
Priority | Units produced | Machine hours per unit | Machine hours used | Throughput |
|---|---|---|---|---|
1. Product B | 10,000 | 0.20 | 2,000 | 10,000 × $80 = $800,000 |
2. Product A | 8,000 | 0.25 | 2,000 | 8,000 × $97 = $776,000 |
3. Product C | 3,333 | 0.30 | 999.9 | 3,333 × $95 = $316,635 |
Total | 4,999.9 | $1,892,635 |
Whole units leave 0.1 machine hour unused. If fractional output were allowed, the remaining 1,000 hours would produce 3,333⅓ units of C.
8. Calculate cost per factory hour and TPAR
Total non-material factory cost in the budget is:
Product A: 8,000 × ($30 labour + $25 variable overhead + $15 fixed overhead) = $560,000
Product B: 10,000 × ($24 + $20 + $12) = $560,000
Product C: 6,000 × ($36 + $30 + $18) = $504,000
Total factory cost = $1,624,000.
Cost per factory hour = $1,624,000 ÷ 5,000 = $324.80
Measure | Product A | Product B | Product C |
|---|---|---|---|
Return per machine hour | $388.00 | $400.00 | $316.67 |
Cost per factory hour | $324.80 | $324.80 | $324.80 |
TPAR | 1.19 | 1.23 | 0.98 |
The old download gave different ratios; the calculations above are corrected. Product B has the strongest TPAR. Product C is below 1, so management should investigate how to improve its throughput rate or reduce its use of the bottleneck.
Profit under the plan
Profit = total throughput − total factory cost
$1,892,635 − $1,624,000 = $268,635
This assumes the $1,624,000 factory cost remains fixed for the period. Do not recalculate it using only the units selected unless the scenario says some operating cost is avoidable.
9. How to improve TPAR
Because TPAR is return per factory hour divided by cost per factory hour, it can be improved by:
increasing selling price where the market permits;
increasing sales of products that earn strong throughput per bottleneck hour;
reducing direct material cost without harming quality or reliability;
reducing bottleneck time per unit through design or process change;
preventing bottleneck downtime, breakdowns and waiting;
moving work from the bottleneck to a non-bottleneck resource;
protecting the bottleneck from defective inputs and rework;
subcontracting some bottleneck work;
adding bottleneck capacity where the benefit exceeds the cost; and
reducing total factory cost without weakening the system.
Some actions improve the formula but not performance. For example, producing more unsold inventory does not create throughput. Making non-bottlenecks “more efficient” may increase queues and inventory without increasing sales.
10. Throughput accounting versus conventional limiting-factor analysis
Feature | Conventional limiting-factor analysis | Throughput accounting |
|---|---|---|
Return measure | Contribution per unit of scarce resource | Throughput per unit of bottleneck resource |
Variable costs | All costs classified as variable | Normally direct material only |
Fixed/operating costs | Deducted after total contribution | Treated as total factory cost and compared through TPAR |
Main emphasis | Optimal short-run product mix | Manage and improve the constraint across the whole system |
Inventory | May be produced subject to the model’s assumptions | Does not generate throughput until sold |
11. CBE method and common mistakes
Calculate capacity required for maximum demand for every potentially scarce resource.
Identify the actual bottleneck by comparing required and available capacity.
Calculate throughput as selling price less direct material only.
Divide by time on the identified bottleneck—not total production time.
Rank products, then allocate bottleneck capacity up to maximum demand.
Calculate total throughput and deduct total factory cost once.
Calculate and interpret TPAR if required.
Recommend practical action using the five focusing steps.
Common mistakes: deducting labour and variable overhead when calculating throughput, ranking by throughput per unit, dividing by the wrong resource time, exceeding maximum demand, deducting factory cost more than once, and assuming a TPAR below 1 automatically proves the product should be discontinued.
12. Continue studying PM
In a nutshell: find the bottleneck, protect every minute of it, rank products by throughput per bottleneck hour, and keep improving until the constraint moves.